The bcd (Benchmarking of CNA Detection) pipeline evaluates the performance
of SOTA tools in CNA profile detection, tumor vs. non-tumor classification,
and subclonal structure identification.
Release notes are at docs/release.rst.
The pipeline depends on some tools and packages listed below:
- R (and Rscript), for extracting expression or probability matrix from output of R packages, such as inferCNV and Numbat.
pip install -U git+https://github.com/hxj5/bcd.gitgit clone https://github.com/hxj5/bcd.git
cd bcd
pip install -U .In either case, if you don't have write permission for your current Python
environment, we suggest creating a separate conda environment
or add --user for your current one.
The full manual is at docs/manual.rst.
Previously, we have an R version of the pipeline.